28 December 2023
Nathan Ashby, senior solutions architect, Cisilion
In turn, largely speaking network architectures have followed the same blueprints built to move 1s and 0s around. While technology has evolved to in networks, it is too common that we build our networks with a core switch, access switches and perhaps a distribution layer depending on the size and the outcomes of the business. But with the complexities of AI, more and more sophisticated apps and hybrid working we need more innovative and intelligent networks to handle and deliver successful outcomes.
So, what does innovation in the network look like?
Before we answer this, it is important to change our perspectives on what we design our networks for. One way to look at is to look at networks in terms of the workforce (people), workload (applications and tools the workforce is trying to access) and workplace (which is where we have seen most disruption in the last few years). We can think of these as our three Ws. Networks need to be able to connect all these elements in different ways to enable the business to operate effectively.
With this in mind, traditionally networks were built around everyone being in the same location with perhaps a little bit of remote access via a VPN or to connect the business to a central DC hosting our business data and applications. As our three W’s changed though this approach does not align with our traditional networks.
Even more so, we’ve moved from keeping the lights on and building reliable infrastructure to business transformation. We cannot use the same mindsets as we once did to ensure the success of our businesses. As such the network’s driving force now is to enable business outcomes which means we need to reinvent the network approach.
With this in mind, connectivity and security are now foundational key outcomes, which must ensure a positive user experience all while helping strive to achieve improved sustainability. With the pandemic, we realised connectivity could be better as the workplace element changed. My workplace is no longer just in the office, it can be at home or in a coffee shop or so forth. We must consider how to integrate home users with our SD-WAN fabric or how to provide the same level of protection to someone sitting in a coffee shop as we do in the office where we may have previously invested heavily in firewalls and security tools such as IPS etc.
Businesses need to be building more dynamic infrastructures with the use of automation and better integration to address these changing needs. The adaptability of the network and the agility we can provide with software-defined fabrics are helping deal with this change, whether a change in the location of the workforce or workload. Automation from a networking infrastructure perspective drives change, reducing manpower hours on certain tasks and reenergising time for other activities.
Automation can vary from Infrastructure-as-a-Code (the managing and provisioning of infrastructure through code instead of through manual processes) to an engineer using simple Python to automate a repetitive task they would have done manually before. The value in automation is in automating regular tasks to save time.
Innovative networks also need better programmability – the ability to interface with business software, for example, a management node for a network solution. It’s about the ability to get data out and squeeze data in. Automation is about time saving, but integration is about pulling data out of one place and putting it in another place either bi-directional or one directional to improve business processes. For example, SIEM integration. Programmability is about context and building a bigger picture too, for example, “I have observed some poor network conditions so I’m going to make some policy changes to improve it.” From a security perspective it could be about providing some additional context between security solutions that are going to deliver a better level of threat protection.
By adding Software-as-a-Solution (SaaS) onto the network it can do more for the business; it is an enabler to drive better business outcomes. For example, integrating foundational security with new security layers that ensure security around home working. The key is to ensure your network is adaptable enough to be able to do use these SaaS solutions.
It’s not always necessary to build a new network to cope with today’s complexities if you already have appropriate compute power available and the flexibility to increase and decrease workloads when needed. What is important is to ensure when you update software or try to integrate new software into the network that it results in a good user experience.
User experience is critical to the success of your network. For example, chat bots are in high demand as the result of Chat GPT and the like, but there is also a lot you can do with simple chat bots. People like language driven results. For example, network engineers want to receive a message alert when there is an issue with the network, and they want to query it back in a chat as that is a more natural way of working and helps fit into their workload. It’s also simple to get up and running quickly and doesn’t require them to code because its already been created elsewhere and is available for wider use.
Containerisation is another technique that allows for network innovation. It is important to network innovation because it provides a consistent runtime environment, enables continuous integration and continuous delivery, and is more efficient than virtual machines. Such as building virtual firewalls within our switches or use of apps running natively our already deployed infrastucture to drive quicker more secure results. Containers use fewer resources and deliver higher utilisation of compute resources which helps with sustainability. It also makes it easier to scale applications up or down as needed, which is essential for modern network architectures.
For network engineers these new networks require a move away from coding to working with scripts which requires a different mindset. If they can focus on the benefit that can be achieved from working in this new way and see the network as its own software platform then they don’t have to reinvent the wheel. Lots of the material is already available off the shelf from vendors and managed service providers and through network communities such as the Cisco DevNet community.
IT teams needs to assess how revolutionising their network with automation and AI will help deliver better business outcomes. Showing ROI is not often easy which is why working with a managed service partner can help demonstrate this.
Once you have a revolutionised network using automation and new tools to drive new outcomes, strong user experience and high levels of security what happens next? Plug it in and off you go, right?
No, issues will still arise even with the best designs in the world whether user error or network outage.
By making sure you have a good assurance plan and appropriate monitoring tools in place you can pick up on any issues quickly. A mature assurance solution can prioritise and grade issues in terms of risk and cost and this is achieved by applying intelligence or machine learning to help interpret the data that the network spits out. This kind of visibility of what is happening across your network is now available to everyone and can help change your response to issues from reactive to proactive. You can start to predict likely problems that can then be addressed before they occur which is the ideal goal of an intelligent network.
As technology evolves, our networks need to work harder and be more intelligent to deliver successful business outcomes. By harnessing automation and containerisation, and ensuring assurance across your network, you will be able to pioneer innovation to ensure continued business success. By doing this we can drive innovation in our networks and in turn ensure the best user and business outcomes, instead of just moving our 0 and 1s around.